It is important to maintain acceptable Quality of Service (QoS) levels for real-time applications in a network. Such applications may include, for example, video streaming and Voice over Internet Protocol (VoIP) telephony. Network monitoring and analysis systems are commonly utilized to ensure that networks are meeting their QoS requirements. Such systems typically collect low-level network measurements, such as delay,jitter and packet loss. See, for example, U.S. Patent Application Publication No. 2005/0053009, entitled “Method and Apparatus for Automatic Determination of Performance Problem Locations in a Network,” which is incorporated by reference herein, and discloses root cause analysis techniques for determining particular problem locations in a network comprising a plurality of endpoint devices.
Certain types of low-level network measurements made by a given network monitoring and analysis system may not be sufficient to allow the system to detect particular quality problems. For example, a particular quality problem experienced by a user on a VoIP call may not be attributable to delay, jitter, packet loss or similar low-level network measurements. Accordingly, a system which relies only on these low-level network measurements may be unable to detect the particular quality problem. Examples of quality problems that may not be reflected in low-level network measurements include problems associated with signaling, noise, echo, robotic voice, etc.
Network users typically have only very limited options for reporting a quality problem. For example, the users may be required to make voice or email contact with a customer support service of the network provider, which can be a time-consuming and frustrating process. A so-called “trouble ticket” may be generated and tracked as a result of such a contact. However, the typical processing of trouble tickets can lead to significant delay before the reported problem is addressed. Also, there may be a certain amount of delay between the time the user experiences the problem and the time the user contacts the customer support service. The network conditions that led to the reported quality problem may be intermittent, transient or otherwise difficult to characterize. Thus, the conventional approach is unable to determine the precise network conditions at or near the time the reported problem was experienced by the user. This makes it very difficult to resolve certain types of reported quality problems, such as those that are not reflected in low-level network measurements.